Photo Annotation Overlay · JP Vlog Style

date: '2026-05-01T07:15:50+08:00'
result: pass
prompt:
  text: |-
    请观察照片中的元素、并为每个物件加上有意义的手绘风注解。请填写照片中的物品（例：披萨、汽水）

    [描写规则]
    - 使用像白色笔画的细线手绘线条
    - 一笔画风格、随性、略带不均匀感
    - 沿着物件外围加上描边轮廓
    - 用箭头或虚线做出视线引导

    [文字规则]
    - 手写风格字体（日系可爱感）
    - 句子简短、像自言自语的小碎念
    - 语气偏日记感、带一点情绪

    [注解生成规则]
    - 饮料 → 味道、温度、心情（例：清爽、微甜、刚刚好）
    - 食物 → 口感、好吃程度（例：松软、超好吃）
    - 空间 → 氛围（例：很放松、喜欢这种感觉）
    - 整体 → 一句总结（例：今天有点幸福~）

    [装饰]
    - 适度加入热气、闪光、爱心、星星、小表情等元素
    - 不要过度装饰，保留空白空间
refs:
  - source: yuan.studio/public/images/yuan_kid_history/raw_frame_001.png
    sha256: null
    error: file not found at write time
    upload_url: >-
      https://personalized-video-refs.tos-cn-beijing.volces.com/4a21e4ba8a91/raw_frame_001.png?X-Tos-Algorithm=TOS4-HMAC-SHA256&X-Tos-Credential=AKLTYjM3ZTcxNjI5ZjFkNGZmYjgyMTNjNWRjZjU4N2IzNjY%2F20260430%2Fcn-beijing%2Ftos%2Frequest&X-Tos-Date=20260430T175741Z&X-Tos-Expires=86400&X-Tos-SignedHeaders=host&X-Tos-Signature=e00dec5d3246089dde91813f6c7914ef6c2b1d22eaca22aff4207093249a22d0
provider:
  id: gpt_image_2
  relay: apimart
  config:
    aspect_ratio: '1:1'
    size: '1:1'
    'n': 1
output:
  path: ./photo_annotation_v1_yuan_kid.png
  bytes: 2235587
  wall_seconds: 47.6
  task_id: task_01KQGASAX7GTPXN0EWHDDR31A5
script: experiments/photo_annotation_test/test_v1_yuan_kid.py
cost_yuan: 0.5
notes: >-
  Photo annotation overlay test · 6th template prompt 范式 (vs identity / mood / personal /
  anti-polish / medium-transform). Tests model's ability to PRESERVE source pixels + ADD doodle
  annotations · without redrawing the photo. Chinese annotation rules · Japanese-cute handwriting
  style · 'don't over-decorate · keep white space' constraint. If PASS · unlocks 小红书 / RED /
  Instagram-vlog viral template (¥0.5/张 · for any food/drink/space photo).